Almost every argument about the anti-defection law in India is conducted at the level of episodes. A group of legislators changes sides, a government falls or survives, and the discussion that follows is about the people involved rather than about the provisions that decided what they were permitted to do. The provisions themselves are short, precise, and rarely stated accurately. A member is not disqualified for changing parties in the loose sense in which that phrase is used in a news report. A member is disqualified for defined conduct set out in the Tenth Schedule of the Constitution, subject to defined exceptions, on the decision of a defined authority whose decision the courts may review on defined grounds. An argument about the law that does not begin there is an argument about politics wearing legal dress.

Stated in terms both sides of the debate would accept, the disagreement is this. One side holds that a legislator is returned by voters who chose a party, its symbol, and its stated programme, so that a legislator who abandons that party after election, or who votes against its direction on the floor, breaks faith with the electorate and may properly be made to lose the seat. The other side holds that a legislator is a representative and not a delegate, that the power to disqualify for disobeying a party direction on every vote converts the member into an instrument of the party leadership, and that a legislature in which no member may vote according to judgment or conscience on ordinary business has surrendered the deliberative function that justifies its existence. Both positions are held by serious constitutional lawyers, both can point to text and to history, and the distance between them is not a distance that better drafting alone can close, because at its centre sits a question about what an elected member is for.

This article takes the controversy seriously by doing the unglamorous work first. It states what the Schedule says, paragraph by paragraph, stripped of characterisation. It reconstructs the political world that produced the Constitution (Fifty-second Amendment) Act, 1985, and the further changes made by the Constitution (Ninety-first Amendment) Act, 2003. It sets out the leading judgment, Kihoto Hollohan v. Zachillhu, with the majority and the minority each given their reasoning, because the minority’s objection to the design has never been answered on its own terms. It then states the case for the law and the case against it at full strength, separates the factual disputes from the value disputes, and defends one conclusion that the arithmetic of the provisions supports on its own: that the law has changed the unit of defection rather than its incidence, raising at each reform the number of legislators who must move together to move without consequence.

What the anti-defection law actually says

The anti-defection law is not a statute in the ordinary sense. It is a Schedule to the Constitution, inserted by the Constitution (Fifty-second Amendment) Act, 1985, which received the assent of the President on 15 February 1985 and was brought into force on 1 March 1985. The same amendment altered four articles of the Constitution so that the Schedule would have something to attach to. It amended Articles 101 and 190, which deal with the vacation of seats in Parliament and in the state legislatures, and it inserted a new clause into each of Articles 102 and 191 providing that a person shall be disqualified for being a member of either House of Parliament, or of a state legislature, if that person is so disqualified under the Tenth Schedule. Defection therefore sits inside the Constitution’s own list of disqualifications, alongside holding an office of profit, being of unsound mind as declared by a competent court, being an undischarged insolvent, and ceasing to be a citizen. It is a constitutional disability, not a criminal offence, and nothing in the Schedule creates a prosecution, a fine, or an imprisonment. The only sanction the Schedule knows is the loss of the seat.

That placement was deliberate, and it is worth pausing on, because it explains both the law’s strength and the difficulty of changing it. An ordinary statute on defection could have been repealed or diluted by a simple majority of a legislature whose members had an interest in diluting it, and it might have been vulnerable to challenge as an infringement of the legislator’s freedom of speech inside the House under Articles 105 and 194. A constitutional amendment required a special majority, placed the rule beyond ordinary legislative revision, and carried the authority of the Constitution itself when the challenge came. The price of that placement is that every defect in the design, including the two that this article argues are structural, can be repaired only by another constitutional amendment commanding the same special majority.

Why was defection made a constitutional disqualification rather than an offence?

Defection was made a constitutional disqualification because the mischief was the holding of a legislative seat on a changed allegiance, and the remedy the drafters wanted was the loss of that seat, imposed through the Constitution’s existing disqualification machinery rather than through a prosecution that would have required proof of motive, a trial, and a criminal standard.

The distinction matters for everything that follows. Because disqualification is a status and not a punishment, the Schedule does not ask why a member moved. It does not require proof that money changed hands, that an office was promised, or that the member acted corruptly. It asks only whether defined conduct occurred: whether the member gave up membership, whether the member voted against a direction, whether an independent member joined a party, whether a nominated member joined one too late. A member who crosses the floor out of genuine conviction and a member who crosses it for a price are treated identically, because the Schedule is indifferent to motive. Supporters of the law regard that indifference as its honesty, since corrupt motive is almost never provable. Critics regard it as its central injustice, since it punishes the conscientious dissenter and the purchased defector with the same instrument and cannot tell them apart.

The Schedule opens, in its first paragraph, with definitions, and two of them do quiet but decisive work. The first is the definition of the legislature party, meaning the group consisting of all the members of a House for the time being belonging to one political party. The second is the definition of the original political party, meaning the party to which a member belongs for the purposes of the disqualification provisions. The whole Schedule is built on the relationship between these two ideas. The political party is the organisation outside the House that sets the member a candidate, allots the symbol, and issues directions. The legislature party is that party’s contingent inside a particular House. Thresholds in the Schedule, for the deleted split and for the surviving merger, are calculated against the legislature party, not against the party’s total membership or its strength in some other House. A party with a large national organisation and a small contingent in one assembly has its merger threshold set by the small contingent, which is one reason the arithmetic of defection varies so sharply from House to House.

The third definitional point is the one most accounts omit. The Schedule applies to members of either House of Parliament and to members of the Legislative Assembly or the Legislative Council of a state. It does not apply to a member of a municipal body or a panchayat, though several states have enacted their own anti-defection provisions for local bodies under state law. It does not apply to a party worker, an office bearer, or a candidate who has not been elected. And it attaches to the member, not to the party: a party may split, merge, dissolve, or change its name without the Schedule having anything to say about the organisation as such. What the Schedule polices is the conduct of individual members of a House, judged against their relationship with the party on whose behalf they sit.

The political world that produced the amendment

The amendment of 1985 was not the product of abstract constitutional reflection. It was the product of a period in which the composition of governments in India was being altered, repeatedly and openly, by legislators changing their allegiance between elections, often in return for office. The phrase that survives from that period is a Hindi expression, Aaya Ram Gaya Ram, meaning Ram has come and Ram has gone, coined in Haryana in 1967 after Gaya Lal, a member of the state Legislative Assembly elected from Hassanpur, changed his party allegiance three times in rapid succession. When Gaya Lal returned to the Congress on one of those occasions, the Congress leader Rao Birendra Singh produced him at a press conference in Chandigarh and announced that Gaya Ram had become Aaya Ram. The joke endured because the conduct it described endured, and because it captured something the electorate found intolerable: that the identity of the government could be decided after the election, by the movement of individuals, in transactions the voter had no part in.

The elections of 1967 had broken the pattern of single-party dominance in several states and produced assemblies in which no party commanded a secure majority. In such assemblies a small number of members held the balance, and their movement from one side to another made and unmade ministries. Governments fell in state after state not because an electorate had withdrawn its confidence but because a handful of legislators had withdrawn theirs, frequently taking ministerial office in the government they had helped to install. Haryana itself passed into President’s rule in 1967 after the assembly’s arithmetic collapsed. The problem was not confined to one state or one party, and it was not confined to the opposition: ruling parties induced defections to manufacture majorities as readily as opposition parties induced them to destroy majorities. Defection was a market in which both sides traded, which is why appeals to party morality accomplished nothing and why the eventual remedy had to be a rule that bound everyone.

Defection was not only a disease of the states. At the Centre, the Janata Party government formed after the elections of 1977 disintegrated from within, and in July 1979 Prime Minister Morarji Desai resigned after defections and withdrawals of support had destroyed his majority, without the government being defeated on the floor. A successor ministry under Charan Singh took office with outside support that was withdrawn within weeks, and it resigned before facing a confidence vote. The sequence demonstrated that the instability produced by shifting allegiances was not a consequence of small assemblies or of inexperienced legislators, since it had occurred in the Lok Sabha, in a Parliament elected on a strong national verdict, within two years of that verdict being delivered. It also demonstrated the feature of the practice that most damaged public confidence, which is that the electorate’s decision could be undone by negotiations to which the electorate was not a party and of which it learned only from the result. When the amendment came six years later, it came with the support of parties that had suffered from defection and of parties that had profited from it, and several members raised during its passage the very question that has dominated criticism since, whether disqualification ought to be decided by the presiding officer of the House at all. The question was raised, considered, and answered in favour of the presiding officer, which is why later criticism is properly directed at a deliberate choice and not at an omission nobody noticed.

The Statement of Objects and Reasons appended to the Constitution (Fifty-second Amendment) Bill, 1985, put the government’s case in a single sentence that has been quoted ever since: the evil of political defections has been a matter of national concern, and if it is not combated, it is likely to undermine the very foundations of democracy and the principles which sustain it. The Bill gave effect to an assurance recorded in the Address of the President to Parliament in January 1985 that the government intended to introduce an anti-defection Bill in that session. The Bill was introduced in the Lok Sabha on 24 January 1985 by the government of Prime Minister Rajiv Gandhi, passed by the Lok Sabha on 30 January 1985 and by the Rajya Sabha on 31 January 1985, and received assent on 15 February 1985. The speed is itself part of the record. A constitutional amendment altering the conditions on which every legislator in the country held a seat moved from introduction to assent in a little over three weeks, which tells the reader how complete the political consensus against defection then was, and also how little time was spent on the design questions that have troubled the law ever since.

What attempts to legislate against defection had failed before 1985?

Two earlier constitutional amendment Bills directed at defection were introduced and failed, one in 1973 and one in 1979, and committees had examined the problem from the late 1960s onward, so the amendment of 1985 was the third attempt at a remedy for a problem that had been formally recognised for well over a decade before it was enacted.

The failure of the earlier attempts is instructive rather than incidental. A committee chaired by the Home Minister, Y. B. Chavan, examined defections at the end of the 1960s, when the practice was at its most conspicuous, and the subject remained under official consideration through the following decade. The Constitution (Thirty-second Amendment) Bill, 1973 and the Constitution (Forty-eighth Amendment) Bill, 1979 each sought to write a disqualification for defection into the Constitution, and neither completed its passage. The political difficulty was obvious: a legislature asked to disqualify its own members for conduct that many of them, or their parties, had benefited from was a legislature with divided incentives. It took a general election that produced a very large majority for the governing party, and a Prime Minister who had made the measure a stated commitment, to carry an amendment that individual legislators might privately have preferred to leave unpassed. The reform record since 1985, which this article traces, repeats the pattern: change has come only when the beneficiaries of the existing arrangement were willing to legislate against an advantage they might one day want.

The grounds of disqualification, paragraph by paragraph

Paragraph 2 of the Tenth Schedule contains the operative law. It provides for four distinct situations in which a member incurs disqualification, and they are not four versions of one idea. Two apply to a member who belongs to a political party, one applies to a member elected as an independent, and one applies to a nominated member. The brief for this article describes them as four grounds, and the findable artifact that follows this section is organised on that basis, but the reader should hold the structure clearly: the first two grounds police loyalty after a party member has taken a seat, while the third and fourth police the acquisition of a party allegiance that the member did not have, or did not declare, at the point of entering the House.

The first ground is that a member of a House belonging to a political party shall be disqualified if the member has voluntarily given up membership of that political party. The language is short and, on its face, simple. It does not say resigned. It does not require a letter, a form, or an announcement. It says voluntarily given up, and the Supreme Court has held, in decisions examined in a later section of this article, that those words carry a wider meaning than resignation and may be satisfied by conduct from which the giving up of membership is to be inferred. The ground is therefore both the most flexible in the Schedule and the most contested, because its reach is fixed by adjudication rather than by text. A member who writes to the party president resigning membership is plainly within it. A member who joins another party is within it. A member who campaigns for another party, who attends its meetings as a participant, who publicly repudiates the party that returned the member to the House, or who joins with members of another party to present a claim to form a government, may be within it, depending on what the presiding officer finds the conduct to establish. Between the plain case and the doubtful case there is a large territory in which the outcome depends on the tribunal’s reading of behaviour, and it is in that territory that most litigation under this ground has been fought.

The second ground is the whip ground, and it is the one that reaches into the daily working of a legislature. A member belonging to a political party is disqualified if the member votes or abstains from voting in the House contrary to any direction issued by the political party to which the member belongs, or by any person or authority authorised by that party to issue a direction, without obtaining prior permission in accordance with the party’s own procedure, and the voting or abstention has not been condoned by the party, or by the authorised person or authority, within fifteen days from the date of the voting or abstention. Every element of that sentence does work. The direction may be issued by the party or by someone the party has authorised, so the legal source of the whip is the political party and not the legislature party acting on its own. The direction may be about any vote. The text contains no limitation to motions of confidence, no limitation to money Bills, no limitation to business on which the survival of the government depends. A direction may be issued on an ordinary motion, on an amendment, on a private member’s Bill, or on a procedural question, and disobedience attracts the same consequence as disobedience on a confidence motion. The member has two escapes and only two: prior permission obtained before the vote, or condonation obtained within fifteen days after it. Both escapes are in the gift of the party. A member who disobeys and is neither permitted beforehand nor forgiven afterwards is disqualified, and the member’s reasons, however principled, are not a defence that the Schedule recognises.

The breadth of the whip ground is the single most criticised feature of the law, and it is important to be precise about what the criticism is and is not. The criticism is not that parties may direct their members how to vote; every parliamentary system expects party cohesion, and a member who finds a party’s position intolerable has the option of leaving it and facing the electorate. The criticism is that the Schedule attaches the loss of the seat to disobedience on any question whatsoever, so that the party’s direction is not merely a political expectation backed by political consequences, such as denial of a ticket at the next election or removal from a party office, but a legal command backed by the destruction of the member’s tenure. The member’s vote in the House ceases, in law, to be the member’s own. It is the party’s vote, cast through the member. Whether that is a defensible model of representation is the value question at the centre of this article, and it is taken up in full below. What can be said as a matter of text is that the drafters of 1985 chose the unlimited version of the whip deliberately, in a period when governments were falling on votes of every description, and that every proposal since to confine the ground to votes on which a government stands or falls has been made and not enacted.

The third ground concerns the independent member. A member elected as an independent candidate, having contested and won without a party, is disqualified if that member joins any political party after the election. The reasoning is parallel to the reasoning behind the first ground: the electorate returned a person who stood outside party allegiance, and the acquisition of an allegiance afterwards defeats the basis on which the seat was won. There is no time limit in this ground and no proportion that can cure it, because a single independent member is the whole of the relevant unit. An independent member may support a government, may vote with a party on every question, may sit with a party’s members by arrangement, and may accept office, without incurring this disqualification, provided the member does not join the party. The line the Schedule draws is the formal line of membership, and it is a line that a politically sophisticated independent member can approach very closely without crossing, which is one of several places where the law’s formality produces results that its purpose would not obviously endorse.

The fourth ground concerns the nominated member, and it is the only ground in paragraph 2 that contains a period of grace. A nominated member of a House who is not a member of any political party when nominated, and who has not become a member of a party before the expiry of six months from the date on which the member takes a seat in the House, is disqualified if that member joins any political party after the expiry of that period. The six months run from the taking of the seat, after compliance with the requirement of making and subscribing an oath or affirmation. A nominated member who joins a party within the six months is safe, and a nominated member who already belonged to a party at the time of nomination is outside the ground altogether, since the Schedule’s concern is with an allegiance acquired after entering the House by a member who entered it without one. The grace period reflects a judgment that a person nominated for distinction in a field such as literature, science, art, or social service ought to be allowed a settled interval in which to decide whether to take a party affiliation for the purpose of legislative work. After that interval, the choice is treated as a defection in substance, because the House and the public were entitled to treat the member’s unattached status as part of the basis of the nomination.

Two further features of paragraph 2 must be fixed before the grounds are left. The first is an explanation attached to the paragraph which provides that a member elected to a House on the ticket of a political party shall be deemed to belong to that party, and a member who is in fact a member of a party shall likewise be treated as belonging to it, so that a member cannot escape the Schedule by disputing the formal character of an affiliation that the election record establishes. The second is the treatment of expulsion. The Schedule does not list expulsion by a party as a ground of disqualification, and it does not say that an expelled member becomes unattached and free. The Supreme Court held in G. Viswanathan v. Hon’ble Speaker, Tamil Nadu Legislative Assembly, decided in 1996, that a member who is expelled by the party that set the member up as a candidate does not thereby cease to be a member of that party for the purposes of the Schedule, and that if such a member then joins another political party, the member will be treated as having voluntarily given up membership of the original party and will incur disqualification. The practical consequence is severe and has been much criticised: a party may expel a troublesome member, and the member remains bound to the expelling party’s directions for the purposes of the Schedule while having no standing within it, unable either to obey a party that has cast the member out or to acquire a new allegiance without losing the seat. That result follows from the reasoning the Court adopted to prevent evasion, since a contrary rule would have let any member obtain a free exit by provoking an expulsion, but it is a result that sits uneasily with the Schedule’s stated foundation in the member’s own voluntary act.

Why does the Schedule not require proof that a defector gained from the move?

The Schedule requires no proof of gain because it is built on conduct rather than on motive, so a member who leaves for office, for money, or for a sincerely held conviction is judged by the same test: whether the defined act of abandonment, disobedience, or late joining is established on the material before the presiding officer.

That design was a response to the evidentiary reality of the period that produced the law. The consideration for a defection was rarely documented. It took the form of a promise, an appointment made after the event, or an understanding that no party to it had an interest in recording. A disqualification that depended on proving the bargain would have failed in almost every case in which it was deserved, and would have converted each petition into an inquiry into private negotiations that the participants could defeat by silence. The drafters chose instead to make the act itself the wrong, on the reasoning recorded in the Statement of Objects and Reasons: that the transfer of allegiance by an elected member is a breach of the faith on which the seat was obtained, whether or not a price can be proved for it. The strength of that choice is that the law can operate at all. Its weakness is the one identified at the start of this article. A conduct test cannot distinguish the member who was bought from the member who was persuaded, and any defence of the Schedule must accept that it disqualifies both, or must explain why the distinction, though real, is one the law is entitled to ignore because the electorate’s interest in the composition it chose does not depend on the mover’s reasons.

How a question of disqualification is framed and decided

A disqualification under the Schedule does not begin on the initiative of the House as a whole. It begins when a question arises, in practice when a member of the House presents a petition to the presiding officer alleging that another member has incurred disqualification on a stated ground and setting out the material relied on. The petition must identify the ground in paragraph 2 and the facts said to satisfy it, because the grounds are distinct and the proof differs between them: a whip petition turns on the issue of a direction, the absence of prior permission, and the expiry of fifteen days without condonation, while a petition on the first ground turns on the conduct from which an abandonment of membership is to be inferred. The presiding officer then proceeds under the rules made for the purpose under paragraph 8, which provide for the petition to be forwarded to the member concerned and to the leader of the member’s legislature party for comments, for a preliminary consideration of whether the petition discloses a case, and for an inquiry, which may be conducted by the presiding officer or referred to a committee of the House for investigation and report, before a decision is recorded.

Three features of that procedure deserve attention because they shape outcomes more than the grounds do. The first is that the member complained against remains a member throughout. No provision of the Schedule suspends a member on the filing of a petition, or on the admission of one, and the member continues to sit, to vote, and, if a minister, to hold office, until a decision of disqualification is pronounced. The second is that the standard of decision is that of a tribunal determining a question of status on evidence, not that of a House dividing on a motion, and the presiding officer is expected to record findings on the elements of the ground rather than a bare conclusion. That expectation is the practical content of the holding in Kihoto Hollohan that the officer acts as a tribunal: it is what makes the resulting order reviewable for perversity or for failure to observe natural justice, since a court can test findings against material but cannot test an unreasoned preference. The third feature is that the Schedule provides no consequence for the member’s constituency in the interval. A seat may be represented for a long period by a member whose title to it is under active challenge on grounds that, if established, mean the member ceased to be entitled to it when the act occurred. The disqualification, when pronounced, is generally treated as operating from the decision rather than as unwinding the votes the member cast in the interval, which is a necessary rule for the stability of proceedings and also a reason the timing of the decision matters so much: votes cast by a member later disqualified have counted in divisions that may have determined the fate of a government or of a Bill.

The consequence of disqualification, once pronounced, is that the member ceases to be a member and the seat falls vacant, to be filled by a by-election. Disqualification under the Schedule is not a bar on future candidature as such. The Supreme Court addressed the point in Shrimanth Balasaheb Patil v. Hon’ble Speaker, Karnataka Legislative Assembly, decided in 2019, where disqualification orders had purported to bar the disqualified members from contesting elections for the remainder of the term of the Assembly. The Court upheld the disqualifications but held that the presiding officer had no power under the Tenth Schedule to impose such a bar, since the Schedule authorises the decision that a member has become disqualified and does not authorise additional disabilities of that kind; the disqualified members were therefore free to seek election again, subject to the separate ministerial and remunerative-post bars inserted in 2003, which themselves fall away on re-election. The decision draws a line that matters for the balance of the scheme. The Schedule’s sanction is the loss of the seat held on the allegiance that was abandoned. It is not exclusion from public life, and an adjudicator who treats it as the latter exceeds the instrument. Critics who regard re-election as too easy an escape and supporters who regard it as the scheme’s democratic corrective are both responding to the same feature: the last word in a defection dispute belongs, by design, to the voters of the constituency, provided the decision is taken in time for them to be asked.

The exceptions: the split that was deleted, the merger that survives, and the Chair

As enacted in 1985, the Schedule contained two exceptions of real political significance and a third, narrower one protecting the presiding officer. The first exception, in paragraph 3, protected a split. It provided that a member of a House shall not be disqualified where the member and any other members of the legislature party constitute a group representing a faction formed by a split in the member’s original political party, and such group consists of not less than one-third of the members of the legislature party. The language required a split in the original political party, not merely a division of opinion among its legislators, and it required the group that resulted to reach one-third of the legislature party’s strength. Within those conditions, a member could leave, form or join another party, vote against the former party’s direction, and take office, all without disqualification. Paragraph 3 was the provision through which the law was most often defeated in practice, and it was deleted in its entirety by the Constitution (Ninety-first Amendment) Act, 2003. Nothing in the Schedule any longer protects a split as such. Accounts that continue to describe the one-third split as an available route are describing a provision that ceased to exist with that amendment, and the error is common enough that the reader should treat any statement of the present law that includes a split exception as unreliable in its other particulars as well.

The second exception, in paragraph 4, protects a merger, and it survives. It provides that a member of a House shall not be disqualified where the member’s original political party merges with another political party and the member claims that the member has become a member of that other party, or of a new political party formed by the merger, or that the member has not accepted the merger and has opted to function as a separate group. The paragraph then sets the condition on which a merger is to be taken as having occurred: a merger of the original political party shall be deemed to have taken place only if, and only if, not less than two-thirds of the members of the legislature party concerned have agreed to the merger. Two distinct requirements are therefore in play, and conflating them is the commonest technical error in public discussion of the provision. The first is a requirement about the party as an organisation, that the original political party merges with another. The second is a requirement about numbers in the House, that two-thirds of the legislature party agree. A group of legislators amounting to two-thirds of a legislature party who simply announce that they have joined another party have not, by that arithmetic alone, satisfied paragraph 4, because the paragraph speaks of the merger of the original party, and the protection is framed around a member whose party has merged. How far the second requirement can substitute for the first in adjudication is a question on which the practice of presiding officers and the observations of courts have not been uniform, and it is the operating record of such claims, rather than their design, that the companion article on Tenth Schedule practice examines. What the design itself says is plain enough: the exception was written for the genuine consolidation of parties, and its threshold was set high so that consolidation, and not departure, would be its occasion.

The third exception, in paragraph 5, protects a member who is elected to preside. It provides that a member of a House shall not be disqualified if, having been elected as the Speaker or the Deputy Speaker of the House of the People or of a state Legislative Assembly, or as the Chairman or the Deputy Chairman of the Council of States or of a state Legislative Council, or as the Chairman of certain specified bodies, the member, by reason of that election, voluntarily gives up membership of the political party to which the member belonged immediately before the election, or, having ceased to hold that office, rejoins the party or becomes a member of another party. The purpose is the dignity and impartiality of the Chair. A presiding officer is expected to stand outside party contest while in office, and a rule that disqualified an officer for surrendering party membership on election to the Chair would have penalised exactly the detachment the office requires. The exception is narrow, personal to the office, and has never been a significant route of evasion. It is still a useful control case for the rest of the Schedule, because it shows the drafters accepting, in one place, that a member may properly be required to loosen a party tie in order to perform a constitutional function, while refusing, everywhere else, to accept conscience, constituency pressure, or disagreement on policy as a reason for doing the same.

The arithmetic of the exceptions, and the claim this article defends

The namable claim of this article is that the anti-defection law changed the unit of defection, not its incidence. Before 1985 an individual legislator could cross the floor and keep the seat. Between 1985 and 2003 a group amounting to one-third of a legislature party could do so together by invoking a split. Since the amendment of 2003, a group amounting to two-thirds of a legislature party may move by invoking a merger. Read in sequence, the three positions describe a single variable being adjusted. Each reform raised the number of legislators who must move together in order to move without consequence. None of them removed the possibility of moving without consequence, because each retained an exception whose function was to legitimise collective movement, and a collective movement that satisfies the exception has the same political effect as the individual movements the law was enacted to stop: it alters the strength of parties in the House, it can install or remove a government, and it does so without any voter being consulted between elections.

The claim is not that the law has achieved nothing. A law that requires two-thirds of a legislature party to agree before its members may move safely has plainly made small defections dangerous and has ended the spectacle of a single member auctioning a vote on which a ministry depended. Retail defection, the defection of one or two members at a time for immediate reward, is the practice the Schedule has most effectively suppressed, and it is not a trivial achievement. The claim is about what the law’s own structure predicts. Where the price of safe movement is a threshold expressed as a proportion of the legislature party, political effort will concentrate on reaching the threshold. Negotiation moves from the individual member to the group, from the single conscience or the single ambition to the organised faction, and the question in any crisis becomes not whether members will move but whether enough of them can be assembled to move under the protection of paragraph 4. The surviving merger exception is, in that precise sense, the split exception at a higher threshold: different in degree, different in the difficulty of the coordination it demands, and not different in kind. The series to which this article belongs has repeatedly found that closing one route in a statute redirects conduct to another rather than ending it. Here the redirection is not a matter of inference from enforcement data. It is visible in the arithmetic of the exceptions themselves, in the distance between one-third and two-thirds.

There is a further consequence of threshold design that is seldom stated. A proportion is not a number, and the practical meaning of two-thirds depends entirely on the size of the legislature party to which it is applied. Two-thirds of a legislature party of sixty is forty members, a coalition that must be assembled, held together, and presented as a merger. Two-thirds of a legislature party of three is two members. In a small assembly, or in a House where a party’s contingent is small, the merger threshold may be reachable by a number of members not much larger than the number who once defected individually. The Schedule’s protection against retail defection is therefore at its weakest exactly where assemblies are smallest and majorities are thinnest, which is where defection does the most institutional damage. A rule expressed as a fraction has an elegance that a rule expressed as a number lacks, but the elegance conceals a variation in substance that the drafters either accepted or did not examine.

A worked illustration shows what the fraction conceals. Take a state assembly in which a party holds sixty seats and the government rests on a majority of eight. Under the law as enacted in 1985, twenty of those sixty members could have left together under paragraph 3, formed or joined another party, and brought the government down, without any of them losing a seat. Under the law after 2003, the same twenty members leaving together would each be disqualified, because a split carries no protection. To move safely they would need forty members and a merger of their original party with another. Take instead a party with six seats in the same assembly, supporting a coalition from outside. Two-thirds of six is four, so four members, together with a merger of their small party into another, satisfy the exception, and four members may be more than enough to change the arithmetic of a closely divided House. The rule is the same rule in both cases. Its protective force differs by an order of magnitude, because it is calibrated to the size of the group defecting from rather than to the size of the majority being destroyed. A threshold designed by reference to the legislature party answers the question how much of a party must move before the movement counts as the party’s own act. It does not answer the question a voter would ask, which is how much of the House may change sides between elections without the electorate being consulted, and on that question the Schedule is silent.

Four grounds, one exception: the findable artifact

The table that follows states the four grounds in paragraph 2 and the surviving merger exception in paragraph 4 in the form a reader needs when a real episode is being reported: what triggers the provision, what must be shown, who decides, what follows, and an illustrative situation on each side of the line. The presiding officer’s separate exemption in paragraph 5 is treated in the text above and in the questions at the end of this article, because its occasion is election to the Chair rather than defection. A reader preparing the subject for examination may find it useful to set the four grounds and the merger exception out as a revision deck with the free online PowerPoint viewer, since the distinctions in the table are precisely the distinctions on which questions in this area turn.

Ground or exception	What triggers it	What must be shown	Who decides	Consequence	A situation where it applies	A situation where it does not
Voluntarily giving up membership, paragraph 2(1)(a)	A member of a party abandons that party	Conduct or a formal act from which the giving up of membership is established, resignation not being required	The Chairman or the Speaker of the House	Disqualification and loss of the seat	A member joins another party and campaigns under its symbol while still sitting on the original party’s benches	A member criticises the party leadership in an interview but remains a member and obeys directions in the House
Voting against a direction, paragraph 2(1)(b)	A member votes or abstains contrary to a party direction	A direction issued by the party or its authorised person, no prior permission, and no condonation within fifteen days	The Chairman or the Speaker of the House	Disqualification and loss of the seat	A member votes against the party’s direction on a motion and the party declines to condone the act within fifteen days	A member obtains the party’s prior permission to abstain, or the party condones the abstention within fifteen days
Independent member joining a party, paragraph 2(2)	A member elected as an independent joins a political party after the election	Election as an independent candidate followed by the joining of a party	The Chairman or the Speaker of the House	Disqualification and loss of the seat	An independent member formally enrolls as a member of a party and sits with its legislature party	An independent member supports a government and votes with a party on every question without joining it
Nominated member joining a party late, paragraph 2(3)	A nominated member joins a party after six months from taking the seat	Nomination without party membership, no party joined within six months of taking the seat, and a joining after that period	The Chairman or the Speaker of the House	Disqualification and loss of the seat	A nominated member who joined no party in the first six months accepts party membership in the second year of the term	A nominated member joins a party in the third month after taking the seat
Merger exception, paragraph 4	The original party merges with another party	A merger of the original political party and agreement to it by not less than two-thirds of the legislature party	The Chairman or the Speaker of the House	No disqualification for members who join the merged party or who opt to function as a separate group	Two parties merge as organisations and two-thirds of the smaller party’s legislature party agree to the merger	A group forming two-thirds of a legislature party announces a departure without any merger of the original party as an organisation

Does the Schedule punish a member for speaking, or only for voting and for membership?

The Schedule punishes two things, voting or abstaining against a party direction and the giving up or acquisition of party membership, and it does not in terms punish speech, so a member may argue against a party’s position in debate without incurring disqualification, provided the member votes as directed and remains a member.

That answer states the text, and the text is what the Supreme Court upheld. When the Schedule was challenged as violating the legislator’s freedom of speech and expression and the privileges conferred by Articles 105 and 194, the majority in Kihoto Hollohan rejected the challenge, reasoning that the provisions were directed at the evil of defection and that the freedom of a member to speak in the House was not destroyed by a rule that governed membership and voting. The distinction is real but it is thinner in practice than it appears on the page. A member who may speak against a Bill and must then vote for it performs a kind of deliberation whose conclusion is fixed in advance, and the knowledge that the conclusion is fixed changes the character of the speech that precedes it. Debate in a legislature does much of its work by persuading members who are free to be persuaded. Where no member on the government side may be persuaded to vote differently, and no member on the opposition side may be persuaded either, the speech that remains is addressed not to the House but to the record, to the press, and to the constituency. Critics of the law, including the minority judges in Kihoto Hollohan on the related question of institutional design, treat that transformation as the deepest cost of the Schedule. Supporters reply that debate in a party system was never a conversation among unattached individuals, that members stood for election as party candidates on a party programme, and that the expectation of cohesion on the vote is the ordinary condition of parliamentary government rather than a distortion of it. Both descriptions are accurate accounts of different things: one describes what happens to the function of speech, the other describes what party government has always required of the vote.

The design choice that determines the law’s operation: the presiding officer decides

Paragraph 6 of the Tenth Schedule provides that if any question arises as to whether a member of a House has become subject to disqualification under the Schedule, the question shall be referred for the decision of the Chairman or, as the case may be, the Speaker of the House, and that decision shall be final. Where the question concerns the Chairman or the Speaker personally, it is decided instead by a member of the House elected by the House for that purpose. Paragraph 6 further provides that proceedings in relation to a disqualification question shall be deemed to be proceedings in Parliament, or in the legislature of a state, within the meaning of Articles 122 and 212, which protect such proceedings from challenge on the ground of irregularity of procedure. Paragraph 8 empowers the Chairman or the Speaker to make rules for giving effect to the Schedule, including rules for the procedure of inquiry, subject to approval and to being laid before the House. And paragraph 7, as enacted, provided that notwithstanding anything in the Constitution, no court shall have any jurisdiction in respect of any matter connected with the disqualification of a member of a House under the Schedule. Paragraph 7 no longer operates, for reasons that are the subject of the next section. Paragraph 6 does, and it is the provision that has determined how the law works in practice more completely than any other.

The choice of the presiding officer as the adjudicator was not the only choice available. The Constitution already contains a model for deciding other disqualifications, in Articles 103 and 192, under which a question whether a member has become subject to a disqualification mentioned in Articles 102(1) or 191(1) is referred to the President or the Governor, who decides in accordance with the opinion of the Election Commission. That model was available in 1985, was familiar, and was not used. The drafters instead placed the decision inside the House, with its own presiding officer, and reinforced the placement with a finality clause and an ouster of the courts. The stated reasons were institutional: that defection is a matter of the internal discipline and composition of the House, that the presiding officer is the guardian of the rights and privileges of the House and of its members, that the officer is placed by convention above party contest, and that a decision taken by an external authority on the membership of the House would trench on legislative autonomy. Those reasons are not negligible, and the majority of the Constitution Bench in Kihoto Hollohan accepted them in terms.

Against them stands a fact about the office in India that no convention has removed. The Speaker of an Indian legislature is ordinarily a member of a political party, is elected to the Chair by the majority in the House, does not resign membership of that party on election in the manner expected of the Speaker of the House of Commons in the United Kingdom, and holds the Chair at the pleasure of the majority that installed the officer. The Speaker’s tenure therefore depends on the continuing support of the majority whose members, and whose opponents, are the subjects of disqualification petitions. A petition against a member of the ruling party or coalition asks the Speaker to reduce the strength of the majority on which the Speaker’s own position rests. A petition against a member of the opposition asks the Speaker to reduce the strength of the Speaker’s opponents. In neither case is the adjudicator detached from the consequence of the decision in the way a tribunal is ordinarily expected to be. That is the structural objection, and it was stated by the minority in Kihoto Hollohan with a clarity that later experience has not weakened: the tenure of the Speaker, who is the authority in the Tenth Schedule to decide the dispute, is dependent on the continuous support of the majority in the House, and therefore the Speaker does not satisfy the requirement of an independent adjudicatory authority, which the minority regarded as an attribute of the democratic system that forms part of the basic structure of the Constitution.

The Schedule also prescribes no time within which the presiding officer must decide. Paragraph 6 sets no period, paragraph 8 leaves procedure to rules made by the presiding officer, and no provision of the Constitution supplies a default. The consequence is that a petition may be filed, admitted, and left undecided for a period long enough to determine the political question it was meant to govern, since a member whose disqualification is pending continues to sit, to vote, and to hold office unless and until a decision is made against that member. Delay in this field is not a neutral administrative failing. It is a decision in favour of the member complained against, taken without reasons and without an order that a court can readily review. The operating record of petitions, the periods for which they have remained pending, and the litigation by which courts have sought to compel decisions belong to the article that owns Tenth Schedule practice, the record of defection decisions in practice, and are not repeated here. What belongs to this article is the design point, which is independent of any single episode: a scheme that makes disqualification depend on the order of an adjudicator who has a political interest in the outcome, and that imposes no time limit on that adjudicator, has built its principal weakness into its own text. Every other feature of the law, the grounds, the exceptions, the thresholds, operates only through that decision, and a reader who understands paragraph 6 understands why identical provisions have produced such different results in different Houses.

For the reader who needs the surrounding field rather than this Schedule alone, the complete guide to election law in India places defection among the other disqualifications and the other mechanisms by which a seat is won, held, and lost, and the two articles are best read as companions: that one maps the field, this one examines the contested provision inside it.

Kihoto Hollohan: what the Constitution Bench held, and what the minority would have held

The validity of the Tenth Schedule was challenged in a group of proceedings that reached a Constitution Bench of five judges of the Supreme Court, and the decision in Kihoto Hollohan v. Zachillhu was delivered on 18 February 1992, with the citation 1992 Supp (2) SCC 651. The bench consisted of M. N. Venkatachaliah, L. M. Sharma, J. S. Verma, K. Jayachandra Reddy, and S. C. Agrawal JJ. The operative conclusions had been pronounced earlier, on 12 November 1991, with reasons to follow. The Court divided three to two. The majority, consisting of Venkatachaliah, Jayachandra Reddy, and Agrawal JJ., upheld the Schedule with one provision struck down. The minority, consisting of Sharma and Verma JJ., would have struck down the Constitution (Fifty-second Amendment) Act, 1985 as a whole. On one question the Court was unanimous, and it is the question on which the structure of the present law depends.

The unanimous holding was that paragraph 7 of the Tenth Schedule, which barred the jurisdiction of all courts in respect of any matter connected with disqualification under the Schedule, effected a change in Articles 136, 226, and 227 of the Constitution, being the provisions that confer jurisdiction on the Supreme Court and on the High Courts, and therefore attracted the proviso to clause (2) of Article 368. That proviso requires that a Bill for a constitutional amendment which makes a change in specified provisions, including those articles, must be ratified by the legislatures of not less than one-half of the states before it is presented to the President for assent. The Fifty-second Amendment Bill had not been so ratified. Paragraph 7 was therefore enacted without a condition precedent that the Constitution required for an amendment of that character. The reason the clause fell is a reason about the amending process rather than about defection, and it is explained in the article on how a constitutional amendment is made and when state ratification is required, whose account of the proviso to Article 368(2) is the necessary background to this part of the judgment.

From that shared starting point the two opinions separated on the question of severability, and the separation decided the fate of the law. The majority held that paragraph 7 was severable from the rest of the Schedule. It was an ancillary provision, standing apart from the substantive scheme of disqualification, and its excision left the remaining provisions complete, workable, and unchanged in their operation, except that the presiding officer’s decision would no longer be insulated from review. Paragraph 7 alone was accordingly struck down, and the rest of the Tenth Schedule survived. The minority held that paragraph 7 was not severable. On their reasoning, the exclusion of judicial review was not an ornament on the scheme but part of the bargain on which the scheme had been enacted, and the failure of the ratification requirement meant that the President’s assent to the Bill was not effective to enact it, so that the whole amendment failed. The difference between the two positions is a difference about what a court may do with a constitutional amendment that is partly invalid, and it is a difference with consequences far beyond this Schedule. The majority’s approach preserved a major constitutional reform by removing the single provision that had not been validly enacted. The minority’s approach treated the amendment as a single act of the constituent power which stood or fell as a whole. Readers who want the doctrine that frames such questions, of which severability in a constitutional amendment is one application, will find it in the account of the basic structure doctrine and the limits of the amending power, which is also where the minority’s second and independent ground belongs.

That second ground was the more fundamental one. The minority held that the Speaker is not an independent adjudicatory authority, that an independent adjudicatory machinery for resolving disputes about the competence of members of a House is an attribute of the democratic system which is a basic feature of the Constitution, and that the choice of the Speaker as sole arbiter therefore took the remaining provisions of the Schedule beyond the amending power. The majority did not accept that reasoning. It held that the Speaker and the Chairman occupy a pivotal position in the scheme of parliamentary democracy, that they are the guardians of the rights and privileges of the House, that they are expected to take far-reaching decisions, and that the vesting of adjudicatory power under the Tenth Schedule in them was not exceptionable. It held further that the provisions of the Schedule were salutary and intended to strengthen the fabric of Indian parliamentary democracy by curbing unprincipled and unethical political defections, that paragraph 2 did not violate the freedom of speech of a member or the provisions of Articles 105 and 194, and that the Schedule did not offend the basic structure of the Constitution.

The majority’s holding on the character of the presiding officer’s function is the part of the judgment with the longest reach. It held that in deciding a question of disqualification under paragraph 6, the Speaker or the Chairman acts as a tribunal, and that the decision is accordingly amenable to judicial review by the Supreme Court under Article 136 and by the High Courts under Articles 226 and 227, notwithstanding the statement in paragraph 6 that the decision shall be final. Statutory finality of that kind, the Court held, does not exclude review on the grounds on which the decisions of tribunals are ordinarily reviewed: violation of a constitutional mandate, mala fides, non-compliance with the rules of natural justice, and perversity. At the same time the Court confined the timing of review. Judicial review is ordinarily available only after the presiding officer has made a decision, and not at an interlocutory stage, so that a member cannot halt a disqualification inquiry by approaching a court in anticipation. The Court recognised a narrow class of exceptions for interlocutory orders that would produce grave, immediate, and irreversible consequences, described in the language of quia timet actions, but the general rule it laid down is that courts do not intervene to prevent a decision; they intervene to correct one, or, as later cases developed, to compel one that has been withheld. The grounds of review must be stated narrowly, because they are often overstated in commentary. A court reviewing a disqualification decision does not retry the question whether the member defected. It does not substitute its own assessment of the evidence. It asks whether the tribunal observed the Constitution, acted in good faith, followed natural justice, and reached a conclusion that was not perverse on the material before it. A decision that is merely debatable, or even wrong on a question of fact, is not for that reason reviewable.

Two things about Kihoto Hollohan should be held together, because commentary usually holds only one of them. The first is that the judgment saved the anti-defection law. Without the majority’s severability holding, the entire Schedule would have fallen in 1992 and the law would have had to be re-enacted, with ratification, in a Parliament whose composition and incentives were very different from those of 1985. The second is that the judgment also identified, in the minority opinion, the objection that has governed criticism of the law ever since, and identified it at the moment of the law’s survival rather than in retrospect. The minority did not predict particular episodes. It made a structural claim: that an adjudicator whose tenure depends on the majority cannot be an independent adjudicator of questions that determine the majority. Nothing in the majority opinion refutes that claim as a description. The majority answered a different question, whether the Constitution forbids the choice, and held that it does not. A reader who keeps the two questions separate understands the judgment, and understands why the argument about the law has continued for three decades without either side conceding: the sides are answering different questions, one about constitutional permissibility and one about institutional design, and both answers can be correct at once.

The deeming provision in paragraph 6 deserves separate notice, because it is the provision through which the Schedule attempted to give the presiding officer’s proceedings the protection that Articles 122 and 212 give to legislative proceedings generally. Those articles bar a court from inquiring into proceedings in Parliament or in a state legislature on the ground of an alleged irregularity of procedure. By declaring proceedings under paragraph 6 to be proceedings in Parliament or in the legislature, the Schedule sought to place disqualification inquiries inside that shelter. The majority in Kihoto Hollohan held that the shelter is not a bar to judicial review of the kind the Court recognised, since review for violation of a constitutional mandate, mala fides, failure to observe natural justice, or perversity is not an inquiry into mere irregularity of procedure. The distinction preserves the ordinary working of the House from litigation about its internal arrangements while denying that protection to an order which deprives a member of a seat. It is an example of the judgment’s general method, which was to sustain the scheme Parliament had enacted while withdrawing from it the immunities that would have placed its operation beyond the Constitution’s ordinary controls, and it explains why the law that survived 1992 is in one respect a different law from the law enacted in 1985: the same grounds, the same exceptions, and the same adjudicator, but an adjudicator whose order must be reasoned, procedurally fair, and defensible before a court.

Voluntarily giving up membership: the ground the courts widened

The first ground in paragraph 2 turns on the words voluntarily given up his membership, and the meaning of those words was settled by the Supreme Court in Ravi S. Naik v. Union of India, decided on 9 February 1994 and reported at 1994 Supp (2) SCC 641. The Court held that the words voluntarily given up his membership are not synonymous with resignation and have a wider connotation, that a person may voluntarily give up membership of a political party even though that person has not tendered a resignation, and that even in the absence of a formal resignation, an inference can be drawn from the conduct of a member that the member has voluntarily given up the membership of the political party to which the member belongs. The appellants in that case were members of the Goa Legislative Assembly whose conduct, in accompanying legislators of another party to the Governor and declaring that they no longer supported their own party, was held capable of supporting the inference the Speaker had drawn.

The holding is defensible on the ground the Court gave for it, which is the prevention of evasion. If a formal resignation were required, the ground would be optional. A member could perform every act of defection, join another party’s councils, vote with it, accept office from it, and defeat a petition by the single expedient of never signing a letter of resignation. A provision against defection that can be avoided by withholding a signature is not a provision against defection. Reading the phrase by reference to conduct closes that route, and it aligns the legal test with the political reality the Schedule addresses, since parties and voters alike treat a member’s conduct, and not the member’s paperwork, as the evidence of allegiance.

The cost of the reading is determinacy, and the cost is real. A test based on inference from conduct gives the adjudicator a discretion that a test based on a document does not, because conduct is ambiguous in a way a resignation letter is not. Attending a meeting may be rebellion, or negotiation, or an attempt to persuade colleagues back. A public statement criticising a party’s leadership may be the prelude to departure or the ordinary currency of internal party argument in a democratic organisation. Silence in the face of a party decision may be dissent or indifference. Where the consequence of an adverse inference is the loss of a seat, the member’s incentive is to avoid any conduct capable of being characterised as abandonment, which includes much of the conduct by which a member might legitimately press for a change of policy or of leadership inside a party. Supporters of the wide reading describe it as the only reading that makes the ground work. Critics describe it as making the ground indeterminate, so that a member cannot know in advance where loyalty ends and defection begins, and must therefore keep a margin of safety that suppresses legitimate internal dissent. Both descriptions capture part of the truth, and the tension between them is a smaller version of the tension in the Schedule as a whole: every device that makes the law harder to evade also makes it harder to predict, and the member who is governed by it bears both effects.

The decision in G. Viswanathan, mentioned earlier in connection with expulsion, belongs to the same line of reasoning and extends it. There the Court treated a member who had been expelled by the party that set the member up as a candidate, and who then joined another political party, as having voluntarily given up membership of the original party. The reasoning preserves the Schedule against a manufactured escape, since a member and a party acting in concert could otherwise arrange an expulsion to free the member. Its consequence for a member expelled against that member’s will, by a leadership the member has opposed, is the harshest result in the case law on the Schedule: the member is bound to a party that has excluded the member, may not join another, and may be disqualified for disobeying the directions of a party in which the member no longer has any place. Later decisions continued to work on the consequences of the scheme rather than on the grounds themselves. In Rajendra Singh Rana v. Swamy Prasad Maurya, decided in 2007 and reported at (2007) 4 SCC 270, the Court held that a presiding officer who fails to exercise the jurisdiction vested by the Schedule, or who accepts a claim without making the finding the Schedule requires, does not thereby place the matter beyond judicial review, since a failure to exercise jurisdiction is itself an error that attracts the reviewing power. In Keisham Meghachandra Singh v. Hon’ble Speaker, Manipur Legislative Assembly, decided on 21 January 2020, a bench of R. F. Nariman, Aniruddha Bose, and V. Ramasubramanian JJ. held that a Speaker acting as a tribunal under the Tenth Schedule is bound to decide disqualification petitions within a reasonable period, that three months from the filing of the petition is the outer limit absent exceptional circumstances for which there is good reason, and that Parliament should consider substituting an independent tribunal, headed by a retired judge of the Supreme Court or a retired Chief Justice of a High Court, for the presiding officer as the adjudicator under the Schedule. That a bench of the Supreme Court should recommend, in terms, that the design upheld in Kihoto Hollohan be reconsidered by the legislature is the clearest indication available that the minority’s structural objection has migrated from dissent into the mainstream of judicial thinking, without the majority holding itself having been disturbed.

The settlement of 2003: what the Ninety-first Amendment changed

The Constitution (Ninety-first Amendment) Act, 2003 was the first and remains the only substantial revision of the anti-defection scheme. The Bill was passed by the Lok Sabha on 16 December 2003 and by the Rajya Sabha on 18 December 2003, received the assent of the President on 1 January 2004, and was published in the Gazette on 2 January 2004. It did four things, and only one of them is usually remembered.

The remembered change is the deletion of paragraph 3 of the Tenth Schedule, which removed the protection for a split. The reasoning recorded in the parliamentary consideration of the Bill was that the split exception had been extensively abused and had become the principal route by which the law was circumvented, because a group of the requisite size could defect together without consequence while an individual acting alone could not. A committee of Parliament chaired by Pranab Mukherjee, which examined the Bill, recorded that the exception granted for a split was being misused and that the promise of personal benefit had influenced defections and produced political horse-trading. Deleting the exception rather than raising its threshold was a deliberate choice, and it is worth noticing what the choice implies. Parliament in 2003 was willing to abolish the one-third route altogether. It was not willing to abolish, or to raise, the two-thirds route in paragraph 4. The distinction between a split and a merger was treated as a distinction of substance, between a faction breaking away and parties consolidating, and not merely as a distinction of size. Whether that distinction bears the weight placed on it is the question this article’s central claim presses, since a two-thirds faction that leaves under the name of a merger produces the same change in the House as a two-thirds faction that leaves under the name of a split.

The second change was a cap on the size of the Council of Ministers. The amendment inserted clause (1A) into Article 75, providing that the total number of Ministers, including the Prime Minister, in the Council of Ministers at the Centre shall not exceed fifteen per cent of the total number of members of the House of the People, and a corresponding clause (1A) into Article 164, providing that the total number of Ministers, including the Chief Minister, in a state shall not exceed fifteen per cent of the total number of members of the Legislative Assembly of that state, with the proviso that the number of Ministers in a state shall not be less than twelve. The connection with defection is direct and was the point of the provision. Ministerial office was the currency in which defections were purchased, and a government that could expand its Council of Ministers without limit could always find a price for the members it needed. Capping the size of the ministry caps the supply of the reward. It also addresses an independent abuse, the jumbo ministry assembled to accommodate factions and allies at public expense, and it carries a deliberate exception in the minimum of twelve, which recognises that a small state cannot conduct its business with a Council of Ministers of two or three, as a strict application of fifteen per cent to a very small assembly would have required. The cap is the most quietly consequential part of the 2003 amendment, and the least discussed, because it operates continuously and without litigation: no petition is filed, no tribunal sits, and no judgment is delivered, yet every government formation since has been conducted inside a limit that did not previously exist.

The third and fourth changes attacked the reward from the other side, by disabling the defector from receiving it. The amendment inserted clause (1B) into Article 75 and a corresponding clause into Article 164, providing that a member of either House of Parliament, or of a state legislature, belonging to any political party who is disqualified for being a member under paragraph 2 of the Tenth Schedule shall also be disqualified for appointment as a Minister, for the period commencing from the date of the disqualification until the date on which the term of office as such member would expire or, where the member contests an election to a House before the expiry of that period, until the date on which the member is declared elected, whichever is earlier. It also inserted Article 361B, providing that a member who is disqualified under paragraph 2 shall be disqualified for holding any remunerative political post for the same period. A remunerative political post is defined by the article to mean an office under the Central or a state government where salary or remuneration is paid out of public revenue, or an office under a body owned or controlled by such a government where salary or remuneration is paid by that body, other than where the salary or remuneration is compensatory in nature. The effect of the two provisions together is that disqualification for defection carries a disability beyond the loss of the seat: the member cannot be consoled with a ministership in another capacity or with the chairmanship of a board or corporation, which had been a standard means of rewarding a defector who had lost, or risked, a seat. The bars are, however, expressly escapable by re-election. A disqualified member who contests the resulting election and is returned by the electorate is restored to full eligibility, and the design thereby reserves the final word to the voter, which supporters of the scheme treat as its democratic safety valve and critics treat as an advantage available chiefly to members with the resources and the local standing to win a second contest in quick succession.

Taken together, the changes of 2003 show a legislature that had diagnosed the problem as one of incentives rather than of definitions. It did not narrow the grounds. It did not restrict the whip. It did not change the adjudicator. It removed the smaller of the two collective exceptions, capped the supply of ministerial rewards, and disabled the disqualified from receiving substitute rewards. That is a coherent programme if the object is to raise the price of defection. It is not a programme directed at the objections of principle that had been made against the law since 1985, and on those objections the amendment of 2003 is silent.

The case for the law, in its strongest form

The case for the anti-defection law rests on a proposition about what an election is. In Indian elections, the overwhelming majority of voters vote for a party, identified by its symbol on the ballot and by the programme it has published, and only secondarily for the individual candidate, who is frequently selected by the party late, is often unknown to most electors in the constituency, and campaigns on the party’s record and the party leader’s appeal. On that description, the seat belongs in a meaningful sense to the mandate, and a member who changes party after election is not exercising a personal freedom but transferring to another owner something that was entrusted for a stated purpose. The member who wishes to change allegiance has an honourable course available, which is to resign the seat and seek a fresh mandate on the new allegiance. What the Schedule forbids is not changing one’s mind. It forbids keeping the seat while changing the allegiance on which the seat was won. Put in those terms, disqualification is not a punishment for opinion but the enforcement of the terms on which the office was obtained, and the analogy is with any position of trust where the holder may leave but may not take the position along.

The second strand of the case is stability, and it is an argument about the conditions of government rather than about morality. A parliamentary executive holds office only while it commands the confidence of the lower House, and in a fragmented assembly that confidence can be destroyed by the movement of a small number of members. Between elections, a government must be able to plan, to budget, to negotiate, and to administer on the assumption that its majority will persist through ordinary political weather. Where individual members may cross the floor at will, no such assumption is available, and the energy of a government is diverted from governing to the continuous maintenance of its own numbers, which is itself a corrupting activity, since a government that must constantly buy loyalty will buy it. The experience before 1985, in which ministries in several states were made and unmade by defections and in which the prospect of office was openly the inducement, is the evidence the supporters of the law rely on, and it is not weak evidence. The majority in Kihoto Hollohan gave this strand its judicial form when it described the Schedule’s provisions as salutary and intended to strengthen the fabric of Indian parliamentary democracy by curbing unprincipled and unethical political defections.

The third strand is the argument from corruption. Defection on the scale practised before 1985 was seldom a matter of conviction, and everyone in public life knew it. It was a transaction, and the consideration was office, or money, or both. A legal order that leaves such transactions lawful because it cannot prove the consideration in each case is not neutral; it licenses the market. The Schedule attacks the market by removing the asset being traded. A member’s allegiance ceases to be saleable in small quantities because a lone member cannot deliver a seat to a buyer without losing it. Supporters add that the alternative remedies proposed, such as relying on the electorate to punish defectors at the next general election, are inadequate in a way the record demonstrates: the interval between elections is long, the defector may be rewarded with office in the interval, and by the time the electorate is consulted the government whose formation the defection produced has already governed for years. On this view, the anti-defection law is best understood not as a restriction on legislators but as an anti-corruption measure of a structural kind, comparable to a rule that a trustee may not sell trust property, and the burden on the individual member’s freedom is the ordinary burden that integrity rules impose.

The fourth strand is the argument from comparative practice and from the character of the party system. Party discipline on votes is not an Indian eccentricity. In the United Kingdom, from which India’s parliamentary forms derive, a member who votes against the party on a matter of confidence brings down the government, and party whips enforce cohesion with sanctions that include withdrawal of the whip, which in substance ends a political career. The difference is that those sanctions are political rather than legal, a distinction the critics of the Indian law treat as decisive and its supporters treat as a distinction about enforcement rather than about substance. Supporters also point to constitutions that go further than India. Article 70 of the Constitution of Bangladesh provides that a member elected as a candidate of a political party vacates the seat if the member resigns from that party or votes in Parliament against it, a provision with no condonation mechanism and no exception for conscience. India’s Schedule, with its prior permission and condonation routes and its merger exception, is by comparison a moderated instrument. That comparison does not settle the argument, but it does resist the suggestion that the Indian law is an aberration explicable only by a local failure of democratic commitment.

The case against the law, in its strongest form

The case against the law begins with the model of the legislator it presupposes. A member of a legislature is elected from a constituency, takes an oath to bear true faith and allegiance to the Constitution, and is given by Articles 105 and 194 a freedom of speech in the House that the Constitution protects with unusual explicitness, including immunity from proceedings in any court in respect of anything said or any vote given in the House. That architecture describes a representative who deliberates. It does not describe a delegate who transmits. The whip ground in paragraph 2 converts the member into a delegate of the party leadership on every question that comes before the House, including questions that have nothing to do with the survival of the government, nothing to do with the programme on which the election was fought, and nothing to do with the interests of the constituency. On an ordinary amendment to an ordinary Bill, a member who has read the clause, heard the debate, and concluded that the clause is badly drafted must still vote for it if so directed, or lose the seat. The member’s judgment is not merely unpersuasive in that situation. It is legally worthless. Critics argue that a legislature composed of members whose judgments are legally worthless is a legislature that performs the forms of deliberation without its substance, and that the long-term consequence is visible in the quality of legislation, since a government that cannot lose a vote has no need to persuade its own side, to answer objections, to accept amendments, or to allow a committee’s scrutiny to alter a Bill it has introduced.

The point about the scrutiny of Bills by parliamentary committees is not incidental to the critique but central to it. Committee scrutiny improves legislation only if the findings of a committee can change outcomes on the floor. Where every member’s vote is commanded, a committee report that criticises a Bill becomes an academic document, because the members who wrote it will vote for the Bill in its unamended form if directed to do so. The anti-defection law therefore does not merely compress dissent in debate. It weakens every other institution of the legislative process whose effectiveness depends on the possibility that a vote might go the other way, and it concentrates power in the executive, since in a parliamentary system the direction of the governing party is in practice issued by or with the approval of the government of the day, so that a provision presented as a restraint on legislators operates as a grant of authority to the executive over the legislature that is supposed to hold it to account.

The second strand of the case against is the adjudicator, already examined in design terms above and restated here as the critics frame it. A law whose consequence is the termination of an elected member’s tenure ought to be administered by an authority that has no interest in the answer. The Schedule entrusts it to an officer elected by the majority, removable by the majority, and ordinarily still a member of a party. The minority in Kihoto Hollohan, Sharma and Verma JJ., made exactly this objection and grounded it in the basic structure. The Election Commission, the Dinesh Goswami Committee on Electoral Reforms, and the National Commission to Review the Working of the Constitution each recommended, in different formulations, that the decision be moved out of the House and taken by the President or the Governor on the advice of the Election Commission, as other disqualifications already are. That the recommendation has been made by bodies of such different composition over such a long period, and has not been adopted, is itself evidence that the defect is not an oversight. Retaining the decision inside the House is advantageous to whoever holds a majority in it, and majorities do not legislate away their own advantages.

The third strand concerns the reach of the first ground. A disqualification that may be inferred from conduct, without a resignation, without joining another party, and without any vote in the House, exposes a member to loss of the seat for the expression of disagreement in forms that a healthy party ought to accommodate. When that ground is combined with the holding in G. Viswanathan, under which an expelled member remains bound to the expelling party, the Schedule can be used not only against defectors but against internal critics, who are the members a party system most needs and least tolerates. The fourth strand is the argument from accountability. A member bound on every vote to the party’s direction cannot be held responsible by constituents for any individual decision, because no decision was the member’s. Accountability is displaced upward to the party leadership, which is not elected by the constituency and cannot be removed by it. The voter who objects to a particular vote is told, truthfully, that the member had no choice. Critics regard that as a corruption of representation more thorough than the corruption the law was enacted to prevent, because it operates on every member on every question, whereas the market in defections operated on a few members on a few occasions.

The fifth strand is empirical and is the one this article’s central claim formalises. The law did not end the transfer of legislative strength between parties. It ended the small transfer and legitimised the large one. Governments have continued to be formed and removed by the movement of legislators, but the movement must be organised at scale, under the cover of a merger, a split before 2003, or a resignation and re-election, rather than performed by an individual in the open. A critic may fairly say that the law has made defection more expensive, more collective, and less visible, and that making a practice more expensive is not the same as making it less common where the prize is a government. Supporters answer that raising the cost and the scale of a practice is a real restraint on it, that the number of occasions on which two-thirds of a legislature party can be assembled is far smaller than the number of occasions on which one member could be persuaded, and that a rule need not be leakproof to be worthwhile. That exchange is a genuine disagreement about evidence and about how much imperfection a law may have before it is judged a failure, and it is not resolvable by pointing to the text.

The member between the party and the constituency

Whose mandate does a legislator carry, the party’s or the constituency’s?

A legislator in India carries both, and the anti-defection law is the point at which the Constitution is made to choose between them, because the member is elected from a territorial constituency as an individual, on a party ticket and symbol, and the Schedule resolves the conflict in favour of the party whenever the two mandates point in different directions.

The constituency side of the claim is not sentimental. A member is the only representative the people of a defined territory have in the legislature, and much of what a member does has nothing to do with party programmes: raising a local grievance, pressing for a road, a hospital, a water supply, objecting to a measure that bears heavily on the district’s principal occupation. On such matters the member’s knowledge is local and the member’s accountability is personal, and a party direction issued from a state or national headquarters may be issued in ignorance of the very facts the member was elected to bring into the room. The party side of the claim is equally concrete. Few electors can name the candidate they voted for a year after the election, but almost all can name the party and the leader, and the organisation that finances the campaign, supplies the workers, and provides the symbol is the party. A member who treats the seat as personal property is treating as an individual asset something that was assembled collectively. Political theorists describe the two models as the delegate and the trustee, and most working legislatures live somewhere between them, with party discipline strong on the programme and weak on the particular. The peculiarity of the Indian settlement is that it is not a settlement between the models arrived at by convention and enforced by political sanction, which is how the United Kingdom manages the same tension. It is a legal rule that resolves every conflict in one direction, on every subject, with the loss of the seat as the sanction, and leaves no space at all for the member’s judgment about the constituency to prevail over a direction. That absoluteness, rather than the existence of discipline, is what the critics of the law are objecting to when they are objecting carefully, and it is what the defenders of the law must defend if they are defending the law as enacted rather than a milder law that might have been.

There is a related consequence that receives less attention than it deserves, concerning the internal life of parties. A legislature party in which no member may vote against a direction is a body in which the only effective arena for disagreement is the party meeting, whose proceedings are private, unrecorded, and controlled by the leadership that issues the directions. Disagreement is therefore displaced from a public forum, where it would be visible to electors and would inform them about the state of opinion inside the party that governs them, into a private forum where it can be managed, deferred, or suppressed without trace. Party leaders acquire, as a by-product of the Schedule, a constitutional instrument for disciplining internal rivals that no party constitution gave them: a member who organises against a leadership inside the party risks conduct from which an abandonment of membership may be inferred, and a member who carries the argument to the floor risks the whip. Supporters of the law reply that internal democracy in parties is a genuine good that must be pursued by other means, through party constitutions and election law, and that the stability of governments cannot be made hostage to it. That reply is coherent. It concedes, though, that the Schedule has effects on the distribution of power within parties that were not among its stated objects, and a full account of the law has to count those effects among its costs.

Which disputes are about facts, and which are about values

Template discipline requires a controversy article to separate the disputes that evidence can settle from the disputes it cannot, and in this subject the separation is unusually clean. The factual questions have answers. Whether the Schedule contains a split exception is a factual question, and the answer is that it did until 2003 and does not any longer. Whether the whip ground is confined to confidence motions is a factual question, and the answer is that it is not. Whether a formal resignation is necessary to attract the first ground is a factual question, and Ravi S. Naik answered it in the negative. Whether paragraph 7 was struck down for want of ratification is a factual question, and the whole bench in Kihoto Hollohan agreed that ratification was required. Whether the Council of Ministers is capped is a factual question, and the answer is fifteen per cent, with a minimum of twelve in a state. Whether two-thirds of a legislature party must agree to a merger for the exception to be available is a factual question, and the answer is in paragraph 4. A very large proportion of published writing on this subject gets at least one of those answers wrong, which is why this article has stated them at length before reaching the argument. Disagreement conducted on false premises about the text is not controversy; it is error.

The value questions are different in kind, and they should be named as value questions so that neither side can claim the authority of fact for what is in truth a conviction. Whether a legislator is primarily the bearer of a party mandate or primarily a representative exercising judgment is a value question. Whether the stability of government is worth a measurable reduction in the deliberative character of the legislature is a value question. Whether it is better that a conscientious dissenter lose a seat than that a purchased defector keep one, where a single rule must govern both, is a value question. Whether the electorate’s interest lies in being able to rely on the composition it chose, or in having members free to respond to events and arguments between elections, is a value question. On each of these, reasonable and informed people divide, and a reader who has been given the text accurately is entitled to divide either way. This article takes a position on the questions that the evidence and the text can support, and declines to disguise the remaining questions as if they had been answered.

The position the evidence supports is the structural one. Whatever the reader concludes about the whip, the proposition that the exceptions have operated as thresholds rather than as prohibitions is established by the text and by the sequence of amendments, and the proposition that the adjudicatory design is the scheme’s weakest element is supported by the reasoning of the minority in Kihoto Hollohan, by the repeated recommendations of the Goswami Committee, the Law Commission, and the National Commission to Review the Working of the Constitution, and by the observation of a bench of the Supreme Court in Keisham Meghachandra Singh that Parliament should consider an independent tribunal in place of the presiding officer. When the judges who administer a scheme, the commissions appointed to review it, and the text of the scheme itself point at the same feature, a conclusion about that feature is not an opinion in the partisan sense. It is the finding the record supports.

The reforms that have been proposed and not enacted

The most frequently urged reform is the confinement of the whip ground to votes on which the government stands or falls. The Dinesh Goswami Committee on Electoral Reforms, reporting in 1990, recommended that disqualification on the ground of defection be limited to two situations: a member voluntarily giving up membership of the party, and a member voting or abstaining contrary to the party’s direction in respect of a motion of confidence or no-confidence. The Law Commission of India, in its 170th Report on the Reform of the Electoral Laws, delivered in 1999, recommended that political parties should issue whips only where the government is in danger, that the provisions exempting splits and mergers be deleted, and that pre-poll electoral fronts be treated as political parties for the purposes of the law, so that the discipline of the Schedule would follow the alliance on whose programme the members were elected. The proposal to confine the whip has a long pedigree, a clear drafting route, and an obvious beneficiary, the ordinary legislator, who is also the person who would have to vote for it in a House whose leadership benefits from the unrestricted version. It has not been enacted. The brief for this article directs that the proposal be presented and noted without endorsement, and it is presented here on those terms, as the reform that the official record returns to most often and that legislatures have most consistently declined to adopt.

The second family of proposals concerns the adjudicator. The Goswami Committee, the Election Commission, and the National Commission to Review the Working of the Constitution, reporting in 2002 under the chairmanship of M. N. Venkatachaliah, each proposed that a question of disqualification for defection be decided by the President or the Governor on the advice of the Election Commission, in the manner of Articles 103 and 192. The National Commission also recommended that a defector be barred from holding a ministerial office or a remunerative political post for the remainder of the term, a recommendation that Parliament adopted in part in 2003, though with the escape of re-election, and that the vote of a defector should not be counted in a motion of confidence. The Halim Committee of 1998, chaired by Hashim Abdul Halim, addressed a different weakness and recommended that the expression voluntarily giving up membership be defined comprehensively, and that restrictions be placed on expelled members, which is the gap that G. Viswanathan exposed. The Supreme Court’s own suggestion in Keisham Meghachandra Singh was for an independent tribunal headed by a retired judge, a proposal that differs from the Election Commission route in that it creates a standing adjudicatory body rather than borrowing an existing constitutional one. The proposals differ in machinery and agree in premise: that the decision should not rest where it rests.

The third family of proposals concerns time. A statutory or constitutional period within which a petition must be decided, with a consequence for default, would remove the largest single source of arbitrary outcomes without disturbing the grounds at all. Courts have supplied a norm in its absence, three months in the formulation of Keisham Meghachandra Singh, but a judicial norm addressed to a presiding officer by way of review is a weaker instrument than a rule in the text, since it depends on a litigant approaching a court and on a court being willing to direct the officer in a particular case. The proposals have in common that none of them requires a judgment about the value questions. A reader who supports the law in its fullest form can still consistently support a time limit and an independent adjudicator, because those reforms go to the administration of rules the reader approves. That is what makes the failure to enact them the more telling: the reforms commanding the widest agreement are not the reforms that divide the two sides of the controversy, and they have stalled anyway.

What is settled, what is contested, and what this article concludes

What is settled can be stated shortly. The Tenth Schedule is valid constitutional law, upheld by a Constitution Bench in 1992 with paragraph 7 struck down for want of ratification by the states. Its grounds are four, its operative exception for collective movement is the merger in paragraph 4 with a two-thirds threshold, the split exception was deleted in 2003, the presiding officer decides subject to judicial review on narrow grounds and ordinarily after a decision, a member’s giving up of membership may be inferred from conduct, the Council of Ministers is capped at fifteen per cent with a minimum of twelve in a state, and a disqualified member is barred from ministerial office and from remunerative political posts until the term expires or the member is re-elected. What is contested on the facts is narrower than the volume of argument suggests: principally, how far the paragraph 4 requirement of a merger of the original party can be satisfied by the agreement of two-thirds of its legislators alone, and how far the conduct-based reading of the first ground should extend. What is contested on values is the whole question of what a legislator is, and that question will not be closed by amendment, by judgment, or by this article.

This article is legal information and analysis about a constitutional scheme, offered in general terms; it is not advice on any individual member’s position, and a person facing a disqualification petition requires advice on the facts of that petition, since the outcome of such proceedings turns on particulars of conduct, direction, and timing that general writing cannot assess.

The conclusion the record supports is the one with which this article began, stated with its qualifications attached. The anti-defection law ended the era in which a single legislator could carry a seat across the floor and keep it, and that was the evil named in its own Statement of Objects and Reasons. It did not end the movement of legislative strength between parties, because at every stage it preserved a collective route and adjusted the number required to use it, from one member, to one-third of a legislature party, to two-thirds. It entrusted the enforcement of the whole scheme to the officer of the House with the most direct political interest in its outcomes, imposed no time limit on that officer, and has resisted for three decades the reforms, recommended by committees, commissions, and a bench of the Supreme Court, that would repair the adjudication without touching the grounds. Whether the law was worth enacting is a question on which the reader has the material to judge. Whether it operates as its text promises is a question the text itself answers, and the answer is that it operates as a threshold, administered by an interested tribunal, against individuals, and in favour of groups large enough to satisfy it.

Frequently Asked Questions

Q: What does the anti defection law prohibit?

The anti-defection law, contained in the Tenth Schedule of the Constitution, prohibits a legislator from keeping a seat after engaging in defined acts of defection. It does not prohibit changing political opinion, criticising a party, or leaving a party and resigning the seat. It attaches disqualification to four kinds of conduct: giving up membership of the party on whose behalf the member sits, voting or abstaining against a party direction without permission or condonation, an independent member joining a party after election, and a nominated member joining a party after six months from taking the seat. A member who satisfies the merger exception is protected.

Q: Which conduct attracts disqualification under the Tenth Schedule?

Paragraph 2 of the Tenth Schedule sets out the conduct. A member belonging to a political party is disqualified on voluntarily giving up membership of that party, or on voting or abstaining from voting in the House contrary to a direction issued by the party or its authorised person, without prior permission and without condonation within fifteen days. An elected independent member is disqualified on joining a political party after the election. A nominated member is disqualified on joining a political party after the expiry of six months from the date of taking the seat. Each ground is decided by the presiding officer of the House.

Q: What counts as voluntarily giving up membership under the Tenth Schedule?

A formal resignation is not required. In Ravi S. Naik v. Union of India, decided in 1994, the Supreme Court held that the words voluntarily given up his membership are not synonymous with resignation and have a wider connotation, and that an inference may be drawn from a member’s conduct that membership has been given up. Joining another party is the clearest case. Conduct such as publicly repudiating the party, campaigning for a rival, or joining with another party’s members to claim the right to form a government may support the inference. Private disagreement, criticism of leadership, or a speech against a party position, without more, does not by itself establish the ground.

Q: What happened to the split exemption?

Paragraph 3 of the Tenth Schedule, as enacted in 1985, protected a split. A member was not disqualified where the member belonged to a group representing a faction formed by a split in the original political party and the group consisted of not less than one-third of the members of the legislature party. The exemption was widely used as the route by which groups defected together without consequence. The Constitution (Ninety-first Amendment) Act, 2003 deleted paragraph 3 altogether. No split exemption exists in the Schedule any longer, and any account that describes the one-third split as an available protection is describing the law as it stood before that amendment.

Q: When does a merger protect a defector?

Paragraph 4 of the Tenth Schedule protects a member where the member’s original political party merges with another political party and the member becomes a member of that other party, or of a new political party formed by the merger, or does not accept the merger and opts to function as a separate group. The protection depends on a condition about numbers: a merger of the original political party is deemed to have taken place only if not less than two-thirds of the members of the legislature party concerned have agreed to the merger. A departure by two-thirds of the legislators, without a merger of the original party as an organisation, is not what the paragraph describes.

Q: Who decides a defection petition?

The Chairman or the Speaker of the House concerned decides. Paragraph 6 of the Tenth Schedule provides that if any question arises as to whether a member has become subject to disqualification under the Schedule, the question shall be referred for the decision of the Chairman or the Speaker, and the decision shall be final. Where the question concerns the presiding officer personally, a member elected by the House for that purpose decides instead. The presiding officer frames rules for the procedure of the inquiry under paragraph 8. In deciding, the officer acts as a tribunal, which is why the decision, although described as final, is open to judicial review on limited grounds.

Q: Can a court review the Speaker’s decision?

Yes, on limited grounds and ordinarily only after the decision is made. In Kihoto Hollohan v. Zachillhu, decided in 1992, the Supreme Court struck down paragraph 7 of the Tenth Schedule, which had barred the jurisdiction of all courts, because it altered Articles 136, 226, and 227 and had not been ratified by the states as Article 368 requires. The Court held that the Speaker acts as a tribunal under paragraph 6, so the decision may be reviewed for violation of a constitutional mandate, mala fides, non-compliance with the rules of natural justice, and perversity. A court does not retry the facts or substitute its own view of whether a defection occurred.

Q: What did the amendment deleting the split exemption change?

The Constitution (Ninety-first Amendment) Act, 2003 deleted paragraph 3 of the Tenth Schedule, removing the protection for a split by one-third of a legislature party. The same amendment did three further things. It capped the Council of Ministers at fifteen per cent of the strength of the Lok Sabha or of a state Legislative Assembly, with a minimum of twelve ministers in a state, through Articles 75(1A) and 164(1A). It barred a member disqualified for defection from appointment as a Minister until the term expired or the member was re-elected. And by inserting Article 361B it barred such a member from holding any remunerative political post for the same period.

Q: What is the difference between the whip ground and the voluntary giving up ground?

They attach to different acts and require different proof. The voluntary giving up ground, in paragraph 2(1)(a), is about the member’s relationship with the party: it asks whether the member has abandoned the party, which may be shown by resignation or inferred from conduct. The whip ground, in paragraph 2(1)(b), is about a single act inside the House: it asks whether the member voted or abstained contrary to a direction issued by the party, without prior permission and without condonation within fifteen days. A member may attract one without the other. Voting against a direction while remaining a party member engages only the whip ground. Leaving the party without voting engages only the first.

Q: Does the whip ground apply to every vote in the House?

Yes. Paragraph 2(1)(b) speaks of a direction issued by the political party in respect of voting in the House, and it contains no limitation to motions of confidence, motions of no-confidence, money Bills, or business on which the survival of the government depends. A direction may be issued on any question, including an ordinary motion, an amendment, or a procedural vote, and disobedience without prior permission or condonation within fifteen days attracts disqualification. That breadth is the most criticised feature of the Schedule. The Dinesh Goswami Committee in 1990 and the Law Commission in its 170th Report of 1999 each proposed confining the whip to votes affecting the government’s survival. Neither proposal has been enacted.

Q: What happens when a member is expelled by their political party?

Expulsion is not listed in the Tenth Schedule as a ground of disqualification, and it does not free the member. In G. Viswanathan v. Hon’ble Speaker, Tamil Nadu Legislative Assembly, decided in 1996, the Supreme Court held that a member expelled by the party that set the member up as a candidate continues to be treated as a member of that party for the purposes of the Schedule, and that if the member then joins another political party, the member is taken to have voluntarily given up membership of the original party and incurs disqualification. The reasoning prevents a staged expulsion being used as an escape route. Its effect on a member expelled for genuine dissent is severe, since the member remains bound to a party that has cast the member out.

Q: Can a nominated member join a political party after taking a seat?

Yes, within a limited period. Paragraph 2(3) of the Tenth Schedule disqualifies a nominated member who is not a member of any political party at the time of nomination only if that member joins a political party after the expiry of six months from the date on which the member takes the seat in the House. Joining a party inside the six months is permitted. A nominated member who already belonged to a political party when nominated is outside the ground altogether. The grace period reflects the position of persons nominated for distinction in fields such as literature, science, art, and social service, who enter the House without a party allegiance and are allowed a settled interval in which to decide whether to acquire one.

Q: Can an independent member join a political party after the election?

No, not while keeping the seat. Paragraph 2(2) of the Tenth Schedule provides that a member elected as an independent candidate shall be disqualified if that member joins any political party after the election. There is no grace period, in contrast to the six months allowed to a nominated member, and no proportion can cure the position, since a single independent member is the whole of the relevant unit. The restriction is on formal membership. An independent member may support a government, vote with a party on every question, and accept office, without incurring this disqualification, provided the member does not join the party. The Schedule draws the line at membership itself rather than at political cooperation.

Q: How does condonation work under the whip ground?

Condonation is one of the two escapes built into paragraph 2(1)(b). A member who votes or abstains contrary to a party direction is not disqualified if the member obtained prior permission in accordance with the party’s procedure, or if the party, or the person or authority authorised by it, condones the voting or abstention within fifteen days from the date of the vote or abstention. The period is short and it runs from the act, not from the filing of any petition. Both escapes are controlled by the party rather than by the member or by the presiding officer. A member whose conduct is neither permitted in advance nor forgiven within fifteen days has no defence under the Schedule based on conscience, constituency pressure, or the merits of the question voted on.

Q: What exemption does Paragraph 5 give to a presiding officer?

Paragraph 5 protects a member who is elected to preside over a House. A member elected as Speaker or Deputy Speaker of the House of the People or of a state Legislative Assembly, or as Chairman or Deputy Chairman of the Council of States or of a state Legislative Council, is not disqualified for voluntarily giving up membership of the party to which the member belonged immediately before the election, where the giving up is by reason of that election. The member is likewise protected on rejoining that party, or joining another, after ceasing to hold the office. The purpose is the dignity and impartiality of the Chair: a presiding officer is expected to stand outside party contest, and the Schedule does not penalise the surrender of party membership that the office invites.

Q: Why was Paragraph 7 of the Tenth Schedule struck down?

Paragraph 7 provided that no court shall have any jurisdiction in respect of any matter connected with the disqualification of a member under the Schedule. In Kihoto Hollohan v. Zachillhu the Supreme Court held unanimously that the paragraph in terms and in effect made a change in Article 136 and in Articles 226 and 227 of the Constitution, which confer jurisdiction on the Supreme Court and the High Courts. Such a change attracts the proviso to Article 368(2), which requires ratification by the legislatures of not less than one-half of the states before the Bill is presented for assent. The Fifty-second Amendment Bill had not been ratified. The majority held paragraph 7 severable and struck it down alone, leaving the rest of the Schedule in force.

Q: What did the minority in Kihoto Hollohan hold?

Sharma and Verma JJ., dissenting, held two things. First, that paragraph 7 was not severable from the rest of the Tenth Schedule, so that its invalidity for want of ratification meant that the President’s assent to the Constitution (Fifty-second Amendment) Act, 1985 was ineffective and the whole amendment failed. Second, that the Speaker is not an independent adjudicatory authority, because the Speaker’s tenure is dependent on the continuous support of the majority in the House, and that an independent adjudicatory machinery for disputes about the competence of members is an attribute of the democratic system which is a basic feature of the Constitution. On that reasoning the choice of the Speaker as sole arbiter itself exceeded the amending power. The majority, Venkatachaliah, Jayachandra Reddy, and Agrawal JJ., rejected both propositions.

Q: What cap did the 91st Amendment place on the Council of Ministers?

Article 75(1A), inserted by the Constitution (Ninety-first Amendment) Act, 2003, provides that the total number of Ministers, including the Prime Minister, in the Council of Ministers shall not exceed fifteen per cent of the total number of members of the House of the People. Article 164(1A) makes the corresponding provision for a state, measured against the total number of members of the Legislative Assembly, with the proviso that the number of Ministers, including the Chief Minister, shall not be less than twelve. The minimum protects small states, where fifteen per cent of a small assembly would leave too few ministers to conduct government business. The cap addresses defection indirectly, by limiting the supply of ministerial office with which defections were rewarded.

Q: Can a member disqualified for defection become a minister?

Not during the period of the bar. Articles 75(1B) and 164(1B), inserted by the Constitution (Ninety-first Amendment) Act, 2003, disqualify a member who is disqualified under paragraph 2 of the Tenth Schedule from appointment as a Minister from the date of disqualification until the date on which the term of office as a member would expire or, where the member contests an election to a House before that date, until the member is declared elected, whichever is earlier. Article 361B imposes a parallel bar on holding any remunerative political post for the same period. Re-election is therefore the route out of the disability: an electorate that returns a disqualified member restores eligibility, which places the final judgment with the voters of the constituency.

Q: What reform confining the whip to confidence votes has been proposed?

The proposal is that the whip ground should operate only where the survival of the government is at stake. The Dinesh Goswami Committee on Electoral Reforms, reporting in 1990, recommended that disqualification be limited to a member voluntarily giving up party membership and to a member voting or abstaining contrary to a direction on a motion of confidence or no-confidence. The Law Commission, in its 170th Report of 1999, recommended that parties issue whips only when the government is in danger, and that the exemptions for splits and mergers be deleted. The National Commission to Review the Working of the Constitution, reporting in 2002, supported moving the decision on disqualification to the President or the Governor on the advice of the Election Commission. None of these proposals has been enacted.